3432 LEONE. ST.PIERRE C.

PRICE
AND CHARLES
(210-212(730mm.),13n20~1.5088-96 (@D 1.512515). C120.7;
mol. wt." 168(182), calcd. for t-butylchlorobenzene 21.03'%,
168; infrared indicates mixture of meta and pura, no ortho;
( b ) 17 g., b. p. 220-222; T Z ~ 1.5084-1.5100,
D C118.970, mol.
wt. 182 (168); calcd. for amylchlorobenzene 19.42%, 182.
(c) 17 g., b.p. 251-254", ~ z . ~ O D 1.5190-1.5237, C1 15.370,
calcd. for di-butyl 15.11%; for butylbutenyll* 15.24%.
Identification.-t-Butylbenzene, llz- and p-t-butyltoluenes,
and the various pentylbenzenes were identified by physical
properties and by reference to available standards and in-
frared spectra. No attempt was made to identify the poly-
alkylated derivatives.
t-Butyl-p-xylene.-This compound was prepared in 11%
yield by alkylation of p-xylene ( 5 moles) with isobutylene
(1 mole) in the presence of nitromethane-aluminum chloride
(100 g./28 9 . ) complex a t 25". The product was worked
up as in procedure l,s and fractionated to separate the t-
butyl derivative, b.p. 217O, 1.5000, dZ040.8738. (Bur-
goyne, et aL.,l4 reported 213-217" (735 mm.), n28*6D1.5003,
(11) Mass spectrograph parent mass main peaks (minor peaks)
based on ClaS isotope.
(12) This fraction rapidly discolors potassium permanganate solu-
tion, indicating olefinic bonds.
(13) For $+butylchlorobenzene; I. P. Tsukervanik. J . Gen. Chem.,
(U.S.S.R.), 8, 1512 (1938).
(14) E. E.Burgoyne, T. J. Klose and D. K. Watson, J . Or#. Ckem.,
ao, 1508 (1955).
Vol. 7s
d29.5~p.60.8757, for this compound prepared in 5.2% yield
via the reaction of t-butyl chloride with the Grignard re-
agent prepared from 2-bromo-p-xylene. ) The infrared
spectrum confirmed 1,2,4-trisubstitution and presence of a t-
butyl side chain, and was identical to that for the C,S-aro-
matic obtained from p-xylene and 2-methyl-2-butene.
These data do not eliminate the 1,.7-dimetliyl-4-t-butyl
isomer which might form by isomerization of p- to m-
xylene followed by alkylation, However, the formation of
this isomer is not considered probable in view of the com-
plete absence of (1) nz-xylene in the recovered xylene, and
of (2) the 1,3-dimethyl-5-t-btltyl isomer in the product.
Furthermore, dealkylation of this material over silica
alumina's (Socony-Vacuum "Chrome Bead") a t 375", 1.8
vol. per vol. of catalyst per hour, yielded a liquid product
consisting of 85% p-xylene, 4 3~ 2 % m-xylene, and only
traces of o-xylene and toluene. The gas evolved contained
78y0 isobutylene and 14.4y0 isobutane. Under the same
conditions with pure p-xylene as the feed, the product con-
tained 90% p-, 4 i 2% m-, no o-xylene and 6 =t270 tolu-
ene.
(15) M.J. Schlatter, paper presented before the Division of Petro-
leum Chemistry of the American Chemical Society, Symposium on
Petrochemicals in the Postwar Years, a t the Chicago Meeting, Sep-
tember, 1953.
ILLINOIS
HARVEY,

[CONTRIBUTION FROM THE DEPARTMENT

OF CHEMISTRY, UNIVERSITY OF NOTREDAME]

The Room Temperature Polymerization of Propylene Oxide

C. PRICE'^
BY LEONE. ST. PIER RE'^ AND CHARLES
RECEIVEDDECEMBER
6, 1955

Polymerization of propylene oxide proceeds readily a t room temperature in the presence of powdered potassium hy-
droxide. The polymer has a molecular weight of 3000-5000 and has a high proportion of unsaturated end-groups, apparently
largely of an allyl ether type, with some propenyl type in the primary polymer. The polymerization appears t o be a specific
surface-catalyzed reaction and a chain-type reaction, not a slow stepwise build-up, since polymer formed a t low conversion
is substantially the same molecular weight as a t higher conversion. Yon-oxidative degradation of the polymer has been
shown t o be catalyzed by toluenesulfonic acid a t 270". The products are volatile, and the presence of propionaldehyde and
dimethyldioxane was established. Degradation with evolution of hydrogen and other gases was observed by electron ir-
radiation.

The present work on the preparation of polypro-
pylene oxide2 through the polymerization of pro-
pylene oxide was undertaken with the hope that a
polymer network based on high molecular weight
material of this structure might have useful elasto-
meric properties. Earlier work, using bases a t
elevated temperatures (100-200") and pressures, 3a,b
had been found t o produce polymer of relatively
low molecular weight, in the neighborhood of 1000
t o 2000. Other work reported since the comple-
tion of this inve~tigation,~howeve:, discloses that
the polymerization of propylene oxide by the use
of ferric chloride or ferric hydroxide catalysis a t
50-150° yields high molecular weight polymer,
which is partially rubbery and partially solid.
( 1 ) (a) General Tire and Rubber Co. Fellow, 1951-1953. Re-
search Laboratories, General Electric Co., Schenectady, N. Y. (h)
John Harrison Laboratory of Chemistry, University of Pennsylvania,
Philadelphia 4, Pa. Presented a t the 128th Meeting of the American
Chemical Society, Minneapolis, Minn., September 12, 1955.
(2) First reported by P. A. Levene and A. Walti, J. Biol. Chcm.. 1 5 ,
325 (1927).
(3) (a) U. S. Patent 2,425,845,August 19, 1947, W. J. Toussaint
and H. R. Fife ( t o Union Carbide and Carbon Corp.); (b) Makoto
Okawara, J . Chcm. SOC.Japan, I n d . Chcm. SCC.,66, 335 (1952).
(4) U. S. Patents 2,706,181and 2,706,182,April 12, 1955, M. E.
Priiitt d n d J. &l,Bagett ( t u Dow Chemical Co.).
This claim was made earlier by Staudinger
although he did not specify the polymerization
procedure.
During the course of our work, it was accidentally
discovered that propylene oxide could be poly-
merized readily by powdered potassium hydroxide
a t room temperature and atmospheric pressure,
which led us to a study of catalysis and tempera-
ture effects on this polymerization. Propylene
oxide was stirred a t room temperature with about
10% by weight of a great variety of inorganic and
organic bases. Of those studied, only powdered
potassium hydroxide gave a reasonably fast poly-
merization under these conditions. I n Table I
are listed some of the catalysts used and their
catalytic effects. In all cases where heterogeneity
existed the catalysts were powdered and then con-
tinuously stirred during polymerization. The reat-
tions were all carried out a t room temperature and
the presence of a non-volatile residue, after heating
a portion of the liquid phase on a steam-bath, was
taken to be a positive criterion of polymeriza-
tion.
(5) H. Staudinger, "Die hochmolekril~.reu organiaclirrl Vcrbin-
dungen." Berlin, J. Sprinycr, 1932, p. 295.
July 20, 1956 ROOMTEMPERATURE
POLYMERIZATION
OF PROPYLENE
OXIDE 3433

I
TABLE 6.27 p in the acid-treated products is taken as evi-
ROOM TEMPERATURE
POLYMERIZATION OF PROPYLENE dence that there was complete hydrolysis of the
OXIDEBY SOME
INORGANIC
BASES propenyl ether with only the allylic ether remaining
Time, in the final product. Attempts to estimate the
Base hr. Results relative quantities of these two unsaturated end-
KOH 51 33% polymerized groups failed to yield reliable data. The presence
NaOCHa 72 No polymer of a double-bonded end-group was further shown
LiOH 72 No polymer by the behavior of the polymer toward 1,5-naph-
NaOH 54 No polymer thalene diisocyanate, before and after oxidation
NaNH2 92 1% of a low polymer with hydrogen peroxide in formic acid. The un-
treated polymer increased in viscosity slightly but
Such bases as Triton B and Triton B ethoxide did not gel when treated with diisocyanate equiva-
also polymerized propylene oxide but in low yields lent to the hydroxyl groups. After oxidation, de-
and to low molecular weights. The polymers pro- signed to convert the end-group double bond to a
duced by the action of powdered potassium hy- glycol, thereby converting a monohydric polymer
droxide on propylene oxide were found t o be in a to trihydric, the polymer formed a n insoluble gel
molecular weight range of 3000-5000 by cryoscopic with an amount of diisocyanate equivalent to the
and end group analysis. hydroxyl groups.
The polymerization rate was shown to be sensi- A likely course for the polymerization process
tive to temperature, but since the monomer boils might be outlined
a t 35" only a small range of temperature was open
to investigation a t atmospheric pressure. Runs /O\
were made a t 33, 20, 0 and -78" using a mixture HOG f CHz-CH + HO-CH,-CHOB
of powdered potassium hydroxide in dry propylene I I
oxide. At 0 and -78" no polymerization could CHr CHI
be noted after three days of stirring. The rates of
polymerization a t 20 and 33" are shown in Tables
111, IV and V. HOCH2CHwwOCHzCH
In connection with questions regarding possible I I/
CH3 CHz 0
mechanisms, measurements of the reduced vis-
cosity of 4% solutions of the polymer in benzene +HOCHCHzOB +-- HOCH#JH-OCHzCH' 'CH,
were made. Tables 111, IV and V also list the I I I I
CH3 CHa CH3 HCCHa
changes in viscosity with percentage conversion of /
monomer. The immediate attainment of a high oe
molecular weight and the relatively small increase The abstraction by the negative oxygen in a
thereafter suggests a chain mechanism for the poly- growing chain of a hydrogen six atoms away could
merization involving slow generation of active terminate a growing chain, producing either the
centers followed by rapid stepwise addition of allyl ether end-group, as indicated, or the propenyl
monomer to form the product polyglycol chains. ether end-group, by attack at the methylene rather
Indeed this very scheme was proposed by Staud- than the methyl group. Presumably the specificity
inger6 in 1932. of solid potassium hydroxide for this polymeriza-
End group analysis of these low-temperature tion is due to the fact that its geometry provides a
polymers indicated a much higher double bond to positive potassium geometrically arranged so as
hydroxyl ratio than was found in polymers ob- to be close to each successive negative oxygen atom.
tained commercially. The latter, presumably pre- Polypropylene oxide in a nitrogen atmosphere
pared a t higher temperatures and pressures,6 exhibit was found to be unstable a t 270°, especially in the
a double bond to hydroxyl ratio of less than 0.02/1, presence of toluenesulfonic acid, which decom-
while for those prepared at room temperature this posed i t largely to volatile products. Propionalde-
ratio is O.G7/1. hyde and dimethyldioxane were identified in the
Evidence that the unsaturation was formed decomposition products.
during the polymerization and not during the isola-
tion may be deduced from the infrared spectra of ( CHa
I 1 -acid A CHa
I
polymer isolated by different procedures. The \-CH-CH~-O], ,CHCHq
infrared spectra suggest the unsaturation to be CHaCHzCHO 4- 0 0
present in two forms prior t o treatment with acid 'CH~CH'
I,
during isolation. One such form would be the
allyl ether A (6.06 cc band) and the other the pro- CH:
penyl ether B (6.27 p band). This degradation was markedly reduced in the
H presence of base and thermal decomposition was
I more extensive in the presence of air than under
R-O-C-C=CHz R-0-CH=CH nitrogen.
I
H H
I I Experimental7
CHa
A B A. Analytical Techniques. 1. Hydroxyl Content by
Phthalation-The method used in this type of determina-
The complete elimination of the absorption at (7) Micro-analyses b y Micro-Tech Laboratories, Skokie, Ill.
All
(6) U. S. Patent 2,448,664, May 30. 1944, H. R . Fife and F. H. melting points are uncorrected. Infrared analyses were run by Mr.
Roberts (tu Uniuu Carbide and Cvrbun Corporation). A. Verbiscar on a Baird spectrophotometer, model AB-1-178.
3434 LEONE. ST.PIERREA N D CHARLES
C. PRICE Vol. 78

TABLE
I1
ROOMTEMPERATURE
POLYMERIZATION OF PROPYLENEOXIDE WITH P O W D E R E D POTASSIUM HYDROXIDE
OH, c=c,
Monomer, KOH, Time, Yield, eq./g. eq./g.
Expt. 6. 6. hr. 6. x 10' x 104 ?SP/LB T121D

2a 200 40 42 160b 3.82 ... ...

...
2b 150 20 64 36.5" 2.96 ... 0.141 ...
2c 150 10 48 ... 4.88 ... ,138 ...
2d 420 10 141 25OC ... 1.35 Eld 1,4509
2e 406 30 42 209' 3.61 .. ,136 1,4509
2f 800 60 220 360' 6.02 1.37 ,161 , . .

4% solution in benzene; c = g. of polymer in 100 g. of solution. Isolated by washing product in ether with hydro-
chloric acid and evaporating t o 150' a t 3 mm. Isolated by shaking an ether solution with Amberlite IR-lZO(H) to remove
base, followed by vacuum evaporation. Cryoscopic molecular weight (benzene), 3440.
tion is that outlined by Siggia.* The sample, weighed into the presence of nothing but recovered propylene oxide in
a small ampoule open at one end, was refluxed for two hours this volqtile fraction.
in 10 ml. of 0.685 iM phthalic anhydride in pyridine. About C. Studies on Rate, Temperature, Degree of Polymeri-
10 ml. of carbon dioxide-free water was then added and the zation and Yield in Potassium Hydroxide Catalyzed Poly-
solution was titrated to a phenolphthalein end-point using merizations. 1. Room Temperature Polymerization of
0.05 to 0.01 iV standard sodium hydroxide. T h e difference Propylene Oxide.-Dried propylene oxide (900 g., 15.5
in titer between a blank and the sample gave a measure of moles) and 60 g. (1.07 moles) of finely powdered technical
the amount of anhydride taken up by the alcohol present. grade potassium hydroxide were stirred a t room tempera-
Phthalation was found to be more sensitive and more repro- ture. Samples were removed at intervals and isolated l > > ~
ducible than acetylation on the propylene oxide polymers. dissolving the polymer in ether, shaking with acidic ion ex-
2 . Determination of Unsaturation.-Several methods change resin and, after filtering, evaporating to 140' a t 0 . 5
were investigated in seeking to establish a satisfactory an- rnm. The polymer product was weighed nnd mdyzctl
alytical technique. .A reliable and reproducible technique viscosimetrically (see Table 111).
was finally arrived at and used in the latter stages of the
work. -4 weighed amount of polymer was dissolved in 10 TABLEI11
ml. of chloroform and 10 nil. of glacial acetic acid. .ifter
cooling in an ice-bath, 10 ml. of bromine in glacial acetic THERATEO F POTASSIUM HYDROXIDE
cATA1,YZEn
acid was pipetted in. Alfterstanding for exactly 2 minutes, P O L Y M E R I Z A T I O N OF P R O P I ' L E N E OXInIi AI' ROIJM
25 nil. of 1056 aqueous potassiuni iodide \vas added and the TEMPERATURE
solution was immediately titrated with standard sodium Time, Yield,
thiosulfate using starch indicator. -4 blank was run a t the hr. 70 v a d C 11200

same time and the unsaturation deterrniued by difference. 12 0 0

3. Viscosimetry.-The viscosity runs were made in an
Ostwald viscosimeter using a 5-ml. sample. The particular 35 15 0,111 1,4508
instrument used had a flow tinie for pure benzene of 76.0 40 15 1.4496
seconds a t 20". 48 17 0.121 1,4605
B. Polymerization Procedures. 1. Polymerizations a t 71.5 19 1,4510
elevated temperatures and pressures, as described in the 132 40 0.123 1,4495
patent literature, were carried out a t temperatures up to
140' and pressures up to 320 p.s.i. using sodium methoxide
and sodium hydroxide as catalysts. The products mere oils 2 . Polymerization of Propylene Oxide a t 33".-To dried
with cryoscopic molecular weights of 900-1300, and with propylene oxide (250 g., 4.3 moles) was added 16.7 g. (0.295
hydroxyl contents of 1 to 1.3 X 10-3 equivalent of hydroxyl mole) of dry, finely powdered potassium hydroxide, and
per gram. stirring was commenced. Samples were removed, dissolved
Polymer 1025, a polypropylene oxide purchased from the in ether, neutralized by shaking with acid ion exchange
Union Carbide and Carbon Corp., was a Tvater-white oil. -In resin, and then filtered and solvent removed. T h e yield ( i f
infrared analysis was obtained; vs,,/c 0.0406, n z o1.4493.
~ polymer and its viscosity were determined (see Table I V ) .
2 . Room Temperature Polymerization.-One or two
samples of propylene oxide, being dried nt room temperature TABLE
IV
over potassium hydroxide, were observed to have undergone THERATE OF POTASSIUM
HYDROXIDE
CATALYZED
polymerization to a polymer of appreciably greater molecu- OF PROPYLESE
POLYMERIZATION OXIDEAT 33"
lar weight (ce. 3000) than the samples from the elevated Time, Yield.
temperatures experiments. Qualitative experiments in- hr. ,o
07 ?lsp/C 7lZuD
dicated that this low temperature polymerization was not
effectively catalyzed by lithium, sodium or barium hydrox- 21 46 0.151 1.4509
ides, sodium methoxide, potassium carbonate, sodatnide or 25 54.2 ,162 1,4510
trimethylbenzylammonium hydroxide or ethoxide (see 27 70.6 ,157 1,4510
Table I ) . 33.5 81.2 ,161 1.4510
d number of samples of polymer were prepared by the
low temperature procedure and are described in Table 11. 57.5 88.4 ,167 1.1510
In all cases, propylene oxide which had been dried over
pellet potassium hydroxide and redistilled was used. The In addition to these measurements, the sample isolated
potassium hydroxide for polymerization was finely- pow- a t 57.5 hours, designated C2a, was found to have a hydroxyl
dered immediately before use. content by phthalation of 2.93 X 10-4 equiv./g. and 1.98 X
3. Isolation of Volatile Fractions from Potassium Hy- 10-4 equiv. C=C/g. was found by analysis using bromine
droxide Catalyzed Polymerization.-Dried propylene oxide in acetic acid. The infrared spectrum exhibited a double
(150 g., 2.58 moles) was stirred a t room temperature with bond absorption at 6.06 p .
powdered potassium hydroxide for 68.7 hours. X portion In addition to the isolation of C2a, two other crude frw-
of the liquid phase, weighing 132 g., was poured off and tions of the 57.5-hour reaction mixture were removed and
distilled. X total of 88.2 g. of volatile material was col- the polymer isolated in a different manner.
lected in a Dry Ice-acetone-bath up to a pot temperature of One fraction, designated C2b, was dissolved in ether and
185" a t 0.5 mm. pressure. On redistillation, all of this washed with aqueous 1 LVsulfuric acid. T h e ether layer \vas
material distilled over at a temperature of 34", indicating subsequently washed with dilute aqueous sodium bicarbon-
a t e solution and then twice with distilled water. The
(8) S. Siggia, "Quantitative Organic Analysis uta Functional ether layer was then dried t o 170" at a pressure of 0.8
Groups," John Wiley and Sons, Inc.,New York, N. Y.,
1949, p. 100. mm. This polymer was almost water-white in appearance
July 20, 1936 ROOMTEMPERATURE
POLYMERIZATION
OF PROPYLENE
OXIDE 3435

as compared to a light brown color for C2a. Hydroxyl and in an ice-bath, the second in an acetone-Dry Ice-bath and
unsaturation analyses were run in the same manner as on the third in a liquid nitrogen bath. The system was
C2a and were found to be 3.04 X lo-' equiv. OH/g. and flushed out with dry nitrogen and the decomposition flas!
1.93 X 10-4 equiv. C=C/g. An infrared analysis was also placed in an oil-bath. After heating for 2.5 hours a t 270 ,
carried out and the spectrum proved to be identical to that the residue in the decomposition flask weighed 1.5 g. and
of polymer C2a. the yellow distillate in the ice-bath weighed 7.5 g.
The third fraction, designated C ~ Cwas, isolated by dis- Distillation of the latter gave two reasonably constant
solving in ether and filtering t o remove catalyst. The boiling fractions: one, b.p. 117-118", 2.5 g., had a refrac-
filtrate was then dried to 170' at 0.9 mm. The prod- tive index, n% 1.4082. The physical constants of this
uct was darker brown than C2a and was subjected to the fraction were similar t o those of dimethyl-1 ,Cdioxane,
same analyses as the other two fractions. The unsatura- which was reported to have a boiling point of 114-117°.1
tion was found to be 2.19 X lo-' equiv. C=C/g. and the Anal. Calcd. for C6HlzO2:C, 62.07; H, 10.34. Found:
hydroxyl content 3.27 X lo-' equiv. OH/g. This latter C, 61.34; H , 10.28.
figure is a corrected value as the polymer was found to con- A second fraction, 0.5 g. distilling a t 139-141°, exhibited
tain 2.11 X 10-6 equiv. KOH/g. The infrared spectrum two peaks in the carbonyl region, one a t 5.85 p and t h e
exhibited two peaks in the double bond region, one a t 6.06 p other a t 5.95 p . Another peak was found in the double
and a stronger one a t 6.27 p . bond region a t 6.08 p . This fraction probably contained
3. Polymerization of Propylene Oxide at 20' .-Dried
propylene oxide (140 g., 2.42 moles) and 9.4 g. (0.167 mole) acrylic acid, boiling point 1$1.9", and possibly some pro-
of dry powdered potassium hydroxide were allowed t o react pionic acid, boiling point 141 .
a t 20" with stirring in exactly the same manner as outlined The material boiling below 117" was converted to deriva-
above for polymerization a t 33'. tives of propionaldehyde. A dimedone derivative was pre-
pared by the method of Shriner and Fuson, and, after being
TABLEV recrystallized, melted at 154-156". The dimedone deriva-
tive of propionaldehyde melts at 1 5 5 O . 9 The 2,4-dinitro-
THERATEOF POTASSIUM
HYDROXIDE
CATALYZED phenylhydrazone was prepared by the method of Shriner
POLYMERIZATION
OF PROPYLENE
OXIDEAT 20' and Fuson and, after two recrystallizations, melted a t 145",
Time, Yield, while the same derivative of propionaldehyde melted a t
hr. % vsdc n% 154'.
8.5 0 4. Oxidative Decomposition of Polypropylene 0xide.-
Oxygen was slowly bubbled through polypropylene oxide 2e
23 3.7 (14 9.) heated in a boiling water-bath. Samples were re-
51.5 32.9 1,4514 moved from the decomposition flask a t intervals and
57.2 48.9 0.149 1,4513 analyzed for peroxide content,s acidity and reduced vis-
72.5 66.2 .154 1,4514 cosity. The acidity was determined by titrating with
standard base and ethanolic solution of the polymer. Ke-
105.5 71.2 .157 1.4515 sults are summarized in Table VII.
4. Attempted polymerization of propylene oxide a t 0'
gave a negligible amount of polymer after 62 hours. At TABLE VI1
-78O, no evidence for polymer could be detected after 89 REACTIONOF POLYPROPYLENEOXIDE WITH OXYGENAT
hours. 100'
D. Polymer Degradation. 1. Determination of Ther- Equiv.
mal Stability.-Polypropylene oxide 2b (1.2 9.) was placed Time, acid/g. Peroxide,
in a small flask which was connected to a mercury manom- hours nPQD x 105 % %P/C
eter. The system was flushed with nitrogen and made gas 0 1.4509 1.99 0.23 0.136
tigh,'. The flask was then placed in a bath and heated to
250 . After two hours at this temperature, a pressure dif- 0.5
2
1.4509
1.4513
3.94
3.21
.31
.98
.142
.128
ference of 0.20 cm. was observable in the manometer.
Heating was continued and the bath temperature raised t o 5 1.4508 13.7 2.34 ,069
300°, whereupon the pressure differential was 1.4 cm. At 9 1.4489 36.5 4.04 ,038
this point the system was opened and on closing the pres-
sure very quickly rose to 1.4 cm. once again. The system 5 . Determination of Gaseous Products upon p-Ray
was once more opened and upon blowing nitrogen gas Irradiation of Liquid Polypropylene Oxide.-Polypropylene
through, vapor could be observed emerging from the outlet oxide 2e was placed in a Pyrex cell equipped with a window
tube. 5 X inch in thickness. The liquid was carefully
2. Investigation of Decomposition at 270°.-Samples of degassed on a vacuum line and the container sealed.
polypropylene oxide 2b were weighed into four flasks, Irradiation was thereupon carriedout for a period of 20
designated a, b, c and d, and heated to 270' for four hours. mjnutes using 1.5-million volt electrons a t a current of 0.4
The results are summarized in Table VI. micro-amp., after which time the gases present "ere col-
lected for analysis. The liquid material was analyzed by
TABLEVI infrared and also viscosimetrically. The reduced viscosity
OF POLYPROPYLENE
DEGRADATION OXIDE AFTER FOUR in benzene was found to be 9.p/c 0.122. Unirradiated 2e
HOURSAT 270' had 0.136.
wt.
The gases were analyzed on the mass spectrometer and the
Wt., g . , loss, following molar percentages obtained:
Sample polymer Under % ?SP/C
a 2.4745 +0.0133 N1 92 (I
HP 63.0 HzO 4.8
toluenesulfonic acid CHI 2.4 CaHa 4.9
b +
0.3434 0.0061 Nz 3.4 0.089 E. Attempted Hydroxylation of the Double Bonds.-
KOH Hydroxylation of the double bonds was attempted by the
C 1.3203 Nz 37.3 ,051 method used by Wittcoff, Roach and Millerloin the oxidation
d 2.1052 Air 23.7 .018 of allyl ether to diglycerol. Polypropylene oxide 2f (50 g.),
containing6.02 X lO-'equiv./g.of hydroxyland 1.37 X 10-4
a Insufficient polymer left for viscosity determination; all equiv./g. of C=C, vsP/c 0.161, was dissolved in 300 g. (6.53
residues were deep brown or black liquids. For the start- moles) of formic acid. With stirring, 90 g. of 30y0 hydrogen
ing polymer 2b, vap/c 0.141 peroxide (0.88 mole) was run in slowly. After a short time
the temperature rose to 45' and, with slight cooling, re-
3. Determination of Products from Acid-catalyzed mained a t that point for three hours. The mixture was
Decomposition.-Polypropylene oxide 2c (10 9.) was placed
in a distilling flask and 0.05 g. of p-toluenesulfonic acid was (9) E. C.Horning and M. G. Horning, J . Org. Chem., 11, 96 (1946).
added. T o the outlet arm of this flask a series of three (10) H. Wittcoff, J. R. Roach and S. E. Miller, THIS JOURNAL,
collecting tubes was connected, the first tube being cooled '71, 2666 (1949).
3436 AND G. FORREST
FLEISCIIACKER
HERMAN WOODS VOl. 7s

TABLE VI11 except for stronger absorption a t 3450 cm.-’ (65% trans-
mission).
MAJOR INFRAREDABSORPTIONPEAKSFOR SOMESAMPLES * Note that the relative intensity of the hydroxyl band
OF POLYPROPYLENEOXIDE AND RELATED COMPOUNDS (3400 cm.-l) is greater in this material, in agreement with
Propylene Polypropylene Polypropylene Dimethyl- analytical titration data. The spectrum was unchanged
oxide oxide 2fo oxide 1025b dioxane after y-ray irradiation.
p, % Y % v, % y, %
cm.-1 tr. cm.’-l tr. cm.-l tr. cm.-l tr.
then placed in the refrigerator overnight after which time
3000 5 3480 77 3400 76 2960 61 the excess peroxide was decomposed using aqueous sodium
1500 21 3360 79 2880 47 2850 63 metabisulfite. The polymer was then taken up in ether and
neutralized with 20% sodium hydroxide. An excess of 40
1450 11 2900 17 1715 88 ml. of 20% sodium hydroxide was added and the solution
1405 6 1450 40 1450 72 1465 79 was refluxed for two hours to saponify any formate esters
1375 21 1375 19 1370 58 1405 83 present. The organic portion was again taken up in 300 ml.
1270 9 1345 40 1340 72 1385 73 of ether and shaken with ion exchange resin to neutralize
any base carried over. The ether solution Was evaporated
1147 21 1300 42 1295 76 1350 82 to a low volume and then taken up in dry benzene and
1135 23 1260 56 1260 79 1275 87 filtered in an effort to break the emulsion which had been
1155 57 formed. The benzene solution was then evaporated to a
1105 15 1080-1120 5 1098 21 1125 58 temperature of 165’ a t a pressure of 0.7 mm., yielding 27.5
g. (55% of starting weight) of polymeric material. Bro-
1023 7 1050 42 1050 75 1085 59 mine analysis indicated the complete absence of unsatura-
1035 64 tion and the hydroxyl equivalent per gram of polymer was
948 7 910 45 910 79 950 73 found by phthalation to have risen to 1.02 X The
895 25 807 70 933 72 reduced viscosity had fallen, however, to 0.0805.
Treatment of this hydroxylated polymer with 1,5-naph-
820-40 7 845 83 thalene diisocyanate produced an insoluble gel, in contrast
747 14 to the original polymer, which increased in viscosity but did
a The sample of polypropylene oxide hydroxylated by
not become insoluble.
treatment with hydrogen peroxide was virtually identical PHILADELPHIA, PENNA.

[CONTRIBUTION FROM THE CHEMICAL LABORATORIES

OF THE UNIVERSITY OF MARYLAND]

BY HERMAN
FLEISCHACKER WOODS
AND G. FORREST
DECEMBER
RECEIVED 14, 1955

have been studied. The 1-,2- and 3-methylhexatrienes give

Methods of synthesis of 1-,2- and 3-methyl-1,3,5-hexatrienes
Diels-Alder adducts with maleic anhydride. The trienes undergo cyclization to methylcyclohexadienes by what is probably
an internal Diels-Alder reaction. The methylcyclohexadienes were partially differentiated from the isomeric trienes by
(1) reaction with maleic anhydride and (2) by aromatization.

The study of the conjugated trienes by this Lab- which undergoes polymerization rapidly in the
oratory has now been extended’ from 1,3,5-hexatri- solid state. It is, however, surprisingly stable in
ene, the simplest of the conjugated trienes, to the the liquid state, showing no marked deterioration
monosubstituted hexatrienes, 1-, 2- and 3-methyl- upon standing for a period of several years in the
hexatrienes, the particular isomers of which (I, I1 absence of special precautions. A second isomeric
and 111) are considered as the open chain analogs of substance is present in the catalytic dehydration
toluene. product of the dienol. We first believed this sub-
CH3 stance was another geometrical isomer of l-methyl-
I 1,3,5-hexatriene but i t was subsequently shown to
be a methylcyclohexadiene (IV). The proportion
of the methylcyclohexadiene increases with increas-
ing temperature of dehydration. Further, passing
I I1 I11 l-methyl-1,3,5-hexatriene over alumina a t elevated
temperatures converted i t to methylcyclohexadiene.
I, or one of its isomers, was readily prepared by These two substances were distinguished from
the following reaction sequence: (1) addition of each other by ultraviolet absorption spectra (Table
the methyl Grignard reagent to 2,4-hexadienal to I), hydrogenation experiments and the Diels-Alder
give 2,4-heptadien-6-01; ( 2 ) catalytic dehydration adducts with maleic anhydride. The latter reac-
of the alcohol in the vapor phase with alumina. 1- tion with l-methyl-l,3,5-hexatriene was interesting,
Methylhexatriene2 is a low melting solid (m.p. 22’) for the addition of the anhydride apparently oc-
(1) G.F. Woods and L. H. Schwartzman, THIS JOURNAL, IO, 3394 curred across the triene a t both the 1,4-and 3,6-p0-
(1948); 11, 1396 (1949); G. F. Woods and S. C. Temin, i b i d . , 7 2 , sitions. This fact was established by dehydro-
139 (1950); G. F. Woods, N. C. Bolgiano and D. E. Duggan, i b i d . ,
1 7 , 1800 (1955).
genation of the adduct, and oxidation with perman-
(2) C. J. Enklar ( C . A . , I , 1169, 1704 (1913)) reported on the ganate to yield the tricarboxylic acid IX and tetra-
dehydration of 2,6-heptadien-4-01 with potassium hydrogen sulfate. carboxylic acid X, Fig. 1. The tricarboxylic acid
He obtained a low melting ( - 14.5O) hydrocarbon (no other data were was isolated by extraction with ether (the tetra-
given), which he states i s possibly 1,3,5-heptatriene. K. H. Ou
( A n n . Chim., 13, 175 (1940)) dehydrated 1,5-heptadien-4-01 at 170° was obtained described as “possibly 1,3,5-heptatriene.’’ He reports
with sodium hydrogen sulfate deposited on pumice. A hydrocarbon P boding point of100-101” r n d ~ z Z 21.4788.
u